HEATING OF INDUCTION MOTORS IN INTERMITTENT DUTY MODES

V. A. Zhadan, Yu. V. Kopylov, D. I. Sannikov

(Presented by the scientific seminar of the departments of electrical machines and general electrical engineering)

For the development of a refined method for the thermal calculation of induction motors in intermittent duty modes (IDM) with frequent starts, it is necessary to carry out experimental studies of the heating of these motors when varying the relative on-time duration \(PV\), the number of starts per hour \(h\), the static torque on the motor shaft \(M_c\), and the coefficient of inertia of the rotating masses \(K_j\), within the limits provided for by GOST 183—66.

The experimental studies were carried out on motors of the 4A series with a shaft-center height of 112 mm, of the basic design—4A112M2, 4A112M4, 4A112M6—and on motors with increased slip, specially intended for operation in modes with frequent starts—4AS112M4, 4AS112M6.

During the manufacture of these motors, up to 76 thermocouples were built into them: 36 in the stator winding, 30 in the frame, and 5 thermocouples in each bearing shield. In each section of the stator winding whose temperature field was measured, 12 thermocouples were embedded: 6 thermocouples in the slot part and 6 thermocouples in the end-winding part (Fig. 1). The sections with thermocouples were distributed uniformly around the circumference of the stator in such a way that they were located in all three phases. In the frame, the thermocouples were arranged uniformly around the circumference and along the length of the frame.

The thermoelectric emf was measured using an R-2/1 semiautomatic potentiometer and an N-39 recording potentiometer. The load of the two-pole motors was produced by means of the same induction machine operating in generator mode with self-excitation, and that of the other motors by means of a mechanical brake [2]. Thermal tests of the motors in IDM with frequent starts were carried out with the motors supplied from a powerful network with unregulated voltage, at nominal load on the shaft and with a coefficient of inertia of the rotating masses unchanged for each motor type size. Since the network voltage may vary during operation of the motor, it was continuously recorded by an N-376 recording voltmeter, and subsequently the losses during starting and during operation were recalculated for the actual voltage.

The general pattern of the temperature distribution of the stator winding for all tested motors in IDM with frequent starts is the same and differs little from the temperature distribution in continuous duty [3, 4].

Since the temperature scatter around the circumference of the stator in any cross section is small—as a rule, it does not exceed 5% [3]—and the magnitude of this scatter is determined mainly by technological deviations, primary attention should be given to analyzing the temperature distribution along the length of the winding. As an example, Figs. 1 and 2 show the temperature distribution along the length of the stator winding of a 4A-112-M-4 motor for different relative on-times and different numbers of starts per hour. Analysis of these dependences shows that, as in continuous-duty operation, the end-winding portions of the stator winding are heated more strongly than the slot portion, owing to the poorer conditions of their cooling; the end-winding portions on the drive side are heated more strongly than the end-winding portions on the fan side, because of the asymmetry of the housing-cooling conditions.

The nonuniformity of the temperature distribution along the length of the stator winding can be assessed, for example, by a coefficient equal to the ratio of the temperature of the hottest point in the end windings on the drive side to the temperature of the least heated point in the slot portion of the stator winding.

The nonuniformity of heating at different on-times changes under the action of two opposing factors: first, when the on-time is reduced, the cooling conditions of the housing and, consequently, of the stator winding as a whole deteriorate,

Fig. 1. Dependence of stator-winding overheating along the length for motor 4A-112M-4 at \(h=\mathrm{var.}\), \(\mathrm{PV}=\mathrm{const}\).

which leads to a reduction in the nonuniformity of its heating; secondly, the conditions of heat removal from the end parts to the internal air become worse, which leads to an increase in nonuniformity. In the 4A-112M-4 motor (Fig. 1), the nonuniformity coefficient of winding heating changes from 1.18 at PV-15% to 1.23 at PV-60%. Thus, with an increase in PV, the nonuniformity of heating of the stator winding increases somewhat owing to the stronger influence of the deterioration of the cooling conditions for the end parts.

Changing the number of motor switchings per hour does not affect the temperature skew along the length of the stator winding (Fig. 2). As the number of switchings per hour increases, the curve of the distribution of stator-winding overheating along its length is shifted upward, as a consequence of the increase in losses, almost parallel to itself.

Fig. 2. Dependence of stator-winding overheating along the length for motor 4A-112M-4 at PV = var, h = const

Since, when PV and h are varied over a wide range, the nonuniformity of heating changes only insignificantly, the thermal state of the motor may, as a first approximation, also be assessed for prolonged-duty operation by the average overheating of the stator winding. Moreover, the amplitude of the temperature oscillations over time in the intermittent-starting mode \((h \geq 30)\) does not exceed \(4^\circ\text{C}\) [2]. In this connection, the table gives the values of the stator-winding overheating \(\Theta_m\), averaged over time and along the length of the winding, and the values of the frame overheating \(\Theta_k\), averaged over time and over the surface area of the frame. Also given here are the total heating losses at rated duty, without allowance for starting losses. When the heating losses in individual modes deviate from the rated values, the average overheating of the stator winding and frame is recalculated in proportion to the change in the heating losses.

Also presented are the average values of the energy \(A_p\) released during one

Table 4

Results of thermal tests of 4A-series motors in intermittent modes

start-up in the stator and rotor windings, with a constant coefficient of inertia of the rotating masses and a constant resisting torque.

As an example, Fig. 3 shows how the average temperature rises of the stator winding and the frame vary at different \(h\) and duty cycles for motors 4A112M4 and 4A112M6.

Fig. 3. Effect of the number of starts on the temperature rise of the stator winding \( \bullet \) and the frame \( \times \) at different duty cycles: — — — for motor 4A-112M-4; —— for motor 4A-112M-6.

With an increase in the number of starts, an additional increase in the average temperature rise of the stator winding and the frame is observed, which grows in direct proportion to \(h\). The slope of the characteristics \(\Theta_m, \Theta_k = f(h)\) is determined primarily by the magnitude of the energy losses per start. Thus, it is possible, on the basis of thermal tests, to determine the energy of the starting losses from the formula

\[ A_p = \sum P \cdot \frac{3600}{h} \cdot \frac{\Delta \Theta}{\Theta_0} \cdot \frac{\mathrm{PV}}{100}, \tag{1} \]

where \(\Theta_0\) is the temperature rise of the stator winding or frame at \(h = 0\);

\[ \Delta \Theta = \Theta_h - \Theta_0 \]

\(\Theta_h\) is the temperature rise of the stator winding or frame at the given number of starts \(h\).

For example, by carrying out calculations of the energy released in the stator and rotor windings during one start, from current oscillograms and by formula (1), we obtain the following results:

motor 4A-112M-4:

from current oscillograms — $A_{\mathrm{p}} = 3350$ J; by formula (1) — $A_{\mathrm{p}} = 4800—5900$ J;

motor 4A-112M-6:

from current oscillograms — $A_{\mathrm{p}} = 1530$ J; by formula (1) — $A_{\mathrm{p}} = 1570—1740$ J.

From this it may be concluded that calculating the energy for one start from current oscillograms gives an underestimated result.

The overheating of the stator winding and the frame increases with increasing duty factor, but its dependence on the duty factor is nonlinear, since with an increase in the duty factor, simultaneously with an increase in the average losses over the cycle, the cooling conditions of the motors improve.

Conclusions

1. The nature of the distribution of overheating of the stator winding along its length changes only slightly when the number of starts and the relative duration of operation are varied.

2. In the mode of frequent starts, the dependence of the overheating of the stator winding and the frame on the number of starts per hour is linear, while its dependence on the relative duration of operation is nonlinear.

3. Asynchronous motors with increased slip are in better heating conditions in modes with frequent starts than motors of the basic design.

References

1. B. T. Artyukhov. Asynchronous Motors under Periodic Load. Kiev, Tekhnika Publishing House, 1972.

2. V. A. Zhadan, Yu. V. Kopylov, D. I. Sannikov. Experimental investigation of asynchronous motors in repetitive short-time operating modes with frequent starts. Proceedings of TPI, vol. 284, Tomsk, TSU Publishing House, 1974.

3. V. A. Zhadan, D. I. Sannikov, R. Ya. Klyain. Investigation of the temperature fields of enclosed, ventilated electric motors. Proceedings of TPI, vol. 172, Tomsk, TSU Publishing House, 1967.

4. V. A. Zhadan, Yu. V. Kopylov, D. I. Sannikov. Investigation of the heating of asynchronous motors of type AO. Proceedings of TPI, vol. 212, Tomsk, TSU Publishing House, 1972.